dsh-linghun
Linghun
End Agent infinite planning: the cognitive loop gives judgments a basis, feedback attribution, and ensures plans reach closure; memory is fully ontologized (v0.4.4+)—three write-side ontology entries (event memory ontology / fact knowledge ontology / rule rule ontology) + ontology projections as fir
The plugin will be installed here. Keep web if you are unsure.
npx -y @deepseek-ai/dsh plugin --profile web add dsh-linghun@0.4.5Description
End Agent infinite planning: the cognitive loop gives judgments a basis, feedback attribution, and ensures plans reach closure; memory is fully ontologized (v0.4.4+)—three write-side ontology entries (event memory ontology / fact knowledge ontology / rule rule ontology) + ontology projections as first-class citizens, with injection sections carrying content in order (knowledge → rules → events → cold storage); scoring identities and responsibilities are built into the architecture (consistent standards, criteria bound to sources, same criteria for the same question, drift lists embedded in the scoring domain ontology); integrates with the linghun-assembler A2A memory management team so sub-agents’ judgment histories can feed back into the main agent. Not a persona card, but judgment— for DeepSeek Harness
Compatibility and provenance
Linghun is published as dsh-linghun and currently resolves to version 0.4.5. The Hub verifies its manifest and preserves the exact installation source for reproducible installs.
- DSH compatibility
- *
- Runtime surfaces
- any
- Release source
- npm
- Registry updated